Clinical outcomes in chronic myeloid leukemia (CML) were revolutionized by development of imatinib (IM) and later generation tyrosine kinase inhibitors (TKIs) that target the Bcr-abl oncogene. Currently, ~55% patients with newly diagnosed CML achieve a major molecular response to TKI treatment (MMR) and 17% a complete molecular response (CMR). Because CML patients survive many years, there is interest in identifying those who can discontinue treatment. This interest is encouraged by the 30% incidence of side effects in during long term TKI therapy and the substantial costs of chronic TKI treatment to patients and healthcare systems. Unfortunately, ~60% of subjects with sustained MMR/CMR relapsed after TKI discontinuation in several clinical trials. Given favorable rates results of TKI re-induction with after relapse, a therapy discontinuation attempt is a part of the National Comprehensive Cancer Network (NCCN) guidelines for patients with prolonged CMR. These discontinuation studies suggest some leukemia stem cells (LSCs) persist in remission. Persistent LSCs do not have mutation or duplication of the BCRABL fusion gene, as is seen with overt TKI resistance, but provide a reservoir of cells susceptible to acquiring such mutations, or those leading to blast crisis (BC). Current clinical tools are inaccurate in predicting the likelihood of relapse vs sustained remission post TKI discontinuation. We developed a murine bone marrow transplantation model to define characteristics associated with successful TKI discontinuation. In this model, primary recipients of Bcr-abl transduced syngeneic bone marrow (in chronic phase; CP) were donors to secondary recipients. Secondary recipients were treated with TKIs + other agents, and mice with an MMR were donors for tertiary recipients. Tertiary recipients were followed without treatment. We found a 64% relapse rate in recipients of bone marrow from mice with IM-induced MMR that did not correlate with number of Bcr-abl+ cells from the treated donors or Bcr-abl transcript copies/cell. In prior studies, we found increased expression of Fap1 (a Fas and Gsk3? inhibitor) contributed to Fas resistance and ?catenin/survivin activity in CML-LSCs. We found that the addition of an inhibitor of Fap1 or survivin to IM treatment prevented relapse in Bcr-abl+ bone marrow recipients. Importantly, no tertiary recipients of bone marrow from IM + Fap1 or survivin inhibitor treated mice relapsed over 24 wks of observation (equivalent to 15+ human years). We found a 50x increase Bcr-abl transcript copies/GFP+ cell in the bone marrow of mice treated with IM vs IM + Fap1 or survivin inhibitor. This reflected differences in bone marrow populations in these mice. In transcriptome analysis of bone marrow from mice treated with TKI + a survivin inhibitor, we identified differences in pathways involved in positive regulation of the innate immune response, NOD-like signaling, cytokine production, and regulation of ribonuclease activity. We hypothesize that identifying pathways which are associated with relapse will permit selection of CML subjects able to safely discontinue treatment. Such pathways may be rationale therapeutic targets to permit more subjects to discontinue treatment, or regain an MMR. We will investigate this through two Aims; Aim 1: Identify molecular markers associated with relapse in CMR-CML patients undergoing a therapy discontinuation attempt. TKI treatment will be discontinued per NCCN guidelines. A clinical trial will correlate molecular markers in the subject?s bone marrow with relapse vs sustained therapy free remission. Aim 2: Define characteristics that predispose to relapse after TKI discontinuation in a murine CML model. We will investigate molecular mechanisms predisposing to relapse after TKI discontinuation in patient derived xenografts or murine bone marrow transplants. Implicated pathways will be targeted in pre-clinical studies. In these studies, we employ a pre-clinical murine model to study TKI discontinuation. Results will be translated to inform studies in human CML subjects. The goal is to identify pathways associated with relapse after TKI discontinuation as prognostic indicators and potential future therapeutic targets.